I linked to the techno-centric description from the US government a few days ago, and both Yigal Chamish and Shawn Callahan rightly complained in the comments that it missed whole aspects of KM that are important to the field. What follows is not so much my be-all-end-all definition, but my thoughts on how I think and talk about knowledge management.
A variety of people are talking about defining knowledge management, and the slow effort at clarifying the Wikipedia entry on KM continues. Much of the discussion revolves around trying to figure out how to define something that touches upon so many different aspects of running an organization.
I'm of two minds in this discussion. One part of me says, "KM isn't about technology, but technology makes KM much more possible." This is the part that connects with the techno-centric definitions. This path takes me into discussions of cool new technologies and the knowledge components therein. I also know that this is a sore area for many people in knowledge management, as this path by itself makes KM seem like information technology wrapped up in buzzwords.
Another part of me says, "KM is about how people and businesses use the knowledge of their people to get things done." Said this way, I generally follow a path of discussion around people sharing what they know and what they are doing. In this mode, I might talk about just how one gets people to share their knowledge and sometimes into a discussion of knowledge itself.
Knowledge comes in many forms, from the books we used in college, to the reference manuals and written policies of business, to written correspondence, to the spoken word, to the skills and expertise that we use. The interesting thing for me - and in how I like to work with clients - is how we go about using this knowledge (and information) to help move the business forward. Do we know what we have? Do we know what value it provides? Do we know what we want to do with it? Do we have the processes and tools in place to make it (more) valuable?